[Biopython-dev] Biopython 1.60 plans and beyond

[Peter Cock]

On Fri, Feb 24, 2012 at 12:02 PM, James Casbon <casbon at gmail.com> wrote:
> Hi Peter et al,
>
> Bit late on this, but...
>
> On 18 February 2012 09:39, Peter Cock <p.j.a.cock at googlemail.com> wrote:
>> VCF format? Variant Call Format - Tiago what's
>> you're impression of work in this area?

Ahem, you're -> your.

> If you think the original license is compatible I'd be happy to fold
> PyVCF into biopython, if it fits.

Excellent.

> Aaron Quinlan is evaluating VCF parsers at the moment (not sure if
> he's on this list), so he can probably give you some good feedback.

Sounds good.

> Some unknowns are:
> 1. using cython/binding to a c library for speed (my preference is
> probably pure python for pypy compatibility)

We've not had any cython dependency so far, but it may be
desirable in the future rather that writing lots of boilerplate code
for calling C libraries. However, I'd hope for a pure Python fall
back for Jython and PyPy etc. I presume Windows would be OK?

> 2. where is BCF going, is that going to be important for a VCF lib?

Not sure.

> 3. There is an optional dependency on pysam, how does that fit with
> biopython?  (other replies in this thread indicate this is already the
> case?)

If it was a run time dependency on pysam that is workable. I'm
unclear if pysam supports Windows, while Python 3 is still pending.
Again, my preference is for a pure Python SAM/BAM library
(and that is doable), possibly as a fallback for compiled code.

> I would also like to know:
> 1. is there existing variant code in biopython that a parser needs
> integration with?

Might this tie in with any of the population genetics code?

> 2. are there other (perhaps more promising) formats that we would like
> parse into the same kind of representation, e.g.
> http://genomebiology.com/2010/11/8/R88/abstract

I don't know.

Peter